ENH: Add isin method to DataFrame.

docs. to be rebased

ENH: Add isin method to DataFrame

Basic tests.

Added method and fixed tests.

ENH: Add ordered argument to df.isin()

Expects a sequence of arrays.

Updated release notes for df.isin()

CLN: cleanup

going to remove ordered argument.

Using a dict for ordered matching. Docs

BUG: fixed subselection length check issues.

Updated release notes for df.isin()

remove merge conflict note

Author: TomAugspurger

doc/source/indexing.rst

@@ -456,6 +456,36 @@ and :ref:`Advanced Indexing <indexing.advanced>` you may select along more than
 
    df2.loc[criterion & (df2['b'] == 'x'),'b':'c']
 
+*New in 0.12.0*
+
+DataFrame also has an ``isin`` method.  When calling ``isin``, pass a set of
+values as either an array or dict.  If values is just an array, ``isin`` returns
+a DataFrame of booleans that is the same shape as the original DataFrame, with Trues
+wherever the element is in the sequence of values.
+
+.. ipython:: python
+
+   df = DataFrame({'vals': [1, 2, 3, 4], 'ids': ['a', 'b', 'f', 'n'],
+                'ids2': ['a', 'n', 'c', 'n']})
+
+   values = ['a', 'b', 1, 3]
+
+   df.isin(values)
+
+Oftentimes you'll want to match certain values with certain columns or rows.
+Just make values a ``dict`` where the key is the row or column, and the value is
+a list of items you want to check for.  Make sure to set axis equal to 0 for
+row-wise or 1 for column-wise matching.
+
+.. ipython:: python
+
+   df = DataFrame({'vals': [1, 2, 3, 4], 'ids': ['a', 'b', 'f', 'n'],
+                   'ids2': ['a', 'n', 'c', 'n']})
+
+   values = {'ids': ['a', 'b'], 'vals': [1, 3]}
+
+   df.isin(values, axis=1)
+
 Where and Masking
 ~~~~~~~~~~~~~~~~~
 

doc/source/release.rst

@@ -54,6 +54,7 @@ pandas 0.12
   - Access to historical Google Finance data in pandas.io.data (:issue:`3814`)
   - DataFrame plotting methods can sample column colors from a Matplotlib
     colormap via the ``colormap`` keyword. (:issue:`3860`)
+  - Added ``isin`` method to DataFrame (:issue:`4211`)
 
 **Improvements to existing features**
 

pandas/core/frame.py

@@ -5481,6 +5481,38 @@ def to_period(self, freq=None, axis=0, copy=True):
 
         return self._constructor(new_data)
 
+    def isin(self, values, axis=None):
+        """
+        Return boolean vector showing whether elements in the DataFrame are
+        exactly contained in the passed sequence of values.
+
+        Parameters
+        ----------
+        values : sequence (array-like) or dict of {label: sequence}.
+        axis : {None, 0, 1}
+            Compute isin row-wise (axis=0) or column-wise (axis=1)
+            Mandatory if values is a dict, ignored otherwise.
+
+        Returns
+        -------
+
+        bools : Series of booleans
+        """
+        if not isinstance(values, dict):
+            return self.applymap(values.__contains__)
+
+        else:
+            from pandas.tools.merge import concat
+            if axis == 1:
+                return concat((self[col].isin(vals) for col, vals in
+                               values.iteritems()), axis=1)
+            elif axis == 0:
+                return concat((self.loc[row].isin(vals) for row, vals in
+                               values.iteritems()), axis=1).T
+            else:
+                raise TypeError('Axis must be "0" or "1" when values is a dict '
+                                'Got "%s" instead.' % str(axis))
+
     #----------------------------------------------------------------------
     # Deprecated stuff
 

pandas/tests/test_frame.py

@@ -10633,6 +10633,56 @@ def _check_f(base, f):
         f = lambda x: x.rename({1: 'foo'}, inplace=True)
         _check_f(data.copy()['c'], f)
 
+    def test_isin(self):
+        # GH #4211
+        df = DataFrame({'vals': [1, 2, 3, 4], 'ids': ['a', 'b', 'f', 'n'],
+                        'ids2': ['a', 'n', 'c', 'n']},
+                        index=['foo', 'bar', 'baz', 'qux'])
+        other = ['a', 'b', 'c']
+        result_none = df[['ids', 'ids2']].isin(other)
+        expected_none = DataFrame({'ids':  [True, True, False, False],
+                                  'ids2': [True, False, True, False]},
+                                  index=['foo', 'bar', 'baz', 'qux'])
+
+        assert_frame_equal(result_none, expected_none)
+
+        # axis = None
+        result_none_full = df.isin(other)
+        expected_none_full = DataFrame({'ids': [True, True, False, False],
+                                        'ids2': [True, False, True, False],
+                                        'vals': [False, False, False, False]},
+                                        index=['foo', 'bar', 'baz', 'qux'])
+
+        assert_frame_equal(result_none_full, expected_none_full)
+
+    def test_isin_dict(self):
+        df = DataFrame({'A': ['a', 'b', 'c', 'd'], 'B': [1, 2, 3, 4],
+                        'C': [1, 5, 7, 8]},
+                        index=['foo', 'bar', 'baz', 'qux'])
+        other = {'A': ('a', 'b'), 'B': (1, 3)}
+        result = df.isin(other, axis=1)
+        expected = DataFrame({'A': [True, True, False, False],
+                              'B': [True, False, True, False]},
+                              index=['foo', 'bar', 'baz', 'qux'])
+        assert_frame_equal(result, expected)
+
+    def test_isin_row(self):
+        df = DataFrame({'A': ['a', 'b', 'c', 'd'], 'B': [1, 2, 3, 4],
+                        'C': [1, 5, 7, 8]},
+                        index=['foo', 'bar', 'baz', 'qux'])
+        ind_other = {'foo': ['a', 1, 1],
+                     'bar': ['d', 2, 1],
+                     'baz': ['nn', 'nn', 'nn']}
+
+        result_ind = df.isin(ind_other, axis=0)
+        expected_ind = DataFrame({'A': [True, False, False],
+                                  'B': [True, True, False],
+                                  'C': [True, False, False]},
+                            index=['foo', 'bar', 'baz']).reindex_like(result_ind)
+
+        assert_frame_equal(result_ind, expected_ind)
+
+        self.assertRaises(TypeError, df.isin, ind_other)
 
 if __name__ == '__main__':
     # unittest.main()